718 HAW THORNE fields were unequal, both among years and harvests. This finding nullified predicting alfalfa production from intermittent harvests during a growing season. Measurements of production are required for all har- vests in a Study period. The amounts of “Sr and of °"Cs associated with alfalfa in each field were computed with the same equation used for computing soil fission products. The same uncertainties apply to the meanvalue as a descrip- tion of the amounts of "Sr and of "Cs with a crop as were given earlier for the radionuclide content of the soils of the fields. Compo- nents of variances are as intricately intertwined between the dependent variable and the independent variables as they were for soils. Again, there is a confounding of the real variation in fission-product distribution in Space with variation introduced by sampling and analytical procedures. Measures of the variability of a material maybe expressed in several forms. In Fig. 8 the mean production of alfalfa is indicated for one field with the dispersion among square-meter specimensillustrated as plus or minus one Standard deviation. As expected, the weight of alfalfa varies with the season of the year and from year to year. In addition, the reliability with which the means can be defined also varies among the harvests. The points of greatest interest in studies of a dairy farm are those where man becomes a consumer of the farm products. Figure 9 illustrates the variation of "cs and Sr per unit volume of milk over a 800 TTT TTT TTT TT PTT TTT TTT TTT TTT TT TTT TTF O 1962 4& 1963 © 1964 MEAN AND ONE ~ 400 stTANDARD = DEVIATION ———» oO t bor — bbe | ua z : 200 O 3 a Ubeitbitielisti tat bet ee JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC ' Fig. 8—-Mean weight of alfalfa samples, with dispersion among specimen weights characterized by one standard deviation, in harvests from one field during three growing seasons.

Select target paragraph3